Warm December

17 Dec: Runoff is 6.34 mm/day

One of the warmest and driest Decembers ever recorded at Murakami, close to the field site. No sign of seasonal snow cover yet, and unusually for December, we even managed to measure the discharge.

New water year brings autumn rains

27 Nov: Runoff is 6.45 mm/day

In the hydrograph below you can clearly see the low flow at the end of September is followed by a major runoff event in early October which raises the baseflow for the rest of the month and into November. This is typical for increasing precipitation and falling ET to lead to higher baseflows into autumn. October monthly runoff (151 mm) is more than three times larger than that for September (46 mm).

September low flows

Very low flow conditions, equivalent to 1.17 mm/day (30 Sep)

Sand and woody debris deposited around the staff gauge

The month of September experienced exceptionally dry conditions this year, as shown by the stage hydrograph below. Only two minor precipitation events can be seen, and long periods of baseflow recession. The monthly runoff was one of the lowest on record for Takiya River, at just 46 mm. The lowest recorded monthly runoff during the period of record (2000-2019) was August 2012 with 35 mm.

Comparing the Kadec and Hobo stage data, it seems the newer Hobo logger is giving more stable data readings. Instability in the Kadec data during low flows could be due to effects of water temperature fluctuations not being compensated for adequately.

July 28 flash flood

Debris around the staff gauge and extreme pool scour to the right are evidence of the 28 July flood

Rainy season was over by the end of July. However, a localized downpour on 28 July gave the largest flood so far this summer, with rainfall intensities of up to 65 mm/h (Takane Amedas), and a daily rainfall total of 149 mm (Miomote Amedas). The hydrograph response was rapid, reaching over 1.6 m at 10.30am one hour after rainfall intensities peaked at 17.5 mm in 10 minutes. In contrast, almost no rain fell in nearby Murakami - only 14 mm for the day.

This flood was followed by a 17 day period without any rain, giving a long flow recession.

The photos below show the large amount of sediment that was deposited at the mouth of the tributary, pushing the flow to the opposite bank and leading to fresh deep scour beneath the bedrock outcrop. Fortunately the stage-rating curve seems to be unaffected.

Rainy season?

So far not so much precipitation in July. End of June saw some moderate rainfall and peak stage of about 1.0m. The hydrograph below shows the data from the Kadec logger which was reset in a new sensor pipe on June 5. During lower flows there is some instability in the stage readings which is not showing in the Hobo stage data. Suspected problems in the Kadec sensor (temperature/atmospheric compensation not functioning?) mean we should consider installing a second Hobo logger as back-up and retire the 19 year old Kadec soon.

Site check after June 18 Niigata/Yamagata earthquake

No visible impact at the gauging station after the recent earthquake of 18 June (22:22). Heavy rains over 16-17 June caused a moderate size flood, with peak stage just less than 1.0m. Some snapshots of the flood taken from the time-lapse camera below:

Staff gauge renewal

Today's main task was to renew the staff gauge installation and sensor pipe for the original sensor/logger system (Druck sensor with Kadec logger housed away from the channel in a logger box). In particular the transducer sensor pipe had been in the river since August 2000, so nearly 19 years! The accumulation of sediment inside the sensor pipe meant that it was extremely difficult to remove the sensor and sensor cable from the pipe. Eventually, by rinsing the inside of the pipe with water and prodding with a long slender branch, we were able to free it. We could have damaged the sensor cable by pulling on it, though a brief test indicated it seemed to be functioning okay. The new sensor pipe does not have holes drilled in it like the old one, so it should not fill with sediment so easily. The holes seem to be unnecessary to correctly measure the water pressure in the river at the end of the sensor pipe.

The conditions were good for our work, with water level very low and stable at 0.306m. The elevation of the top of the staff gauge was levelled against a benchmark (100m), and the new staff gauge installed at the same height (99.641m). The two vertical pipes in the river were maintained as they were not easily removed and appear to be very secure (changed previously in 2008). The secondary backup sensor/logger system (Onset Hobo) was reinstalled in the same position behind the staff gauge, snug beneath the pipe clamp, as in the photo below.

Future work is needed to adjust the amount of slack in the sensor cable (spare coils located in the logger box), and to protect the cable hose where it exits the sensor pipe to give it support against snow loading in the winter (see below).

The following two photos show the location of the bedrock benchmark, marked by a white pebble.

The hydrograph below shows the tail end of the snowmelt season during very dry conditions with only two minor rainfall events. See how the diurnal snowmelt signal (sine wave pattern) gradually fades out by the end of May.

April 25 rain-on-snow

A period of fine weather melt with building diurnal snowmelt peaks turned to rain on April 25 to give the sharp rain-on-snow peak in the hydrograph above. Daily precipitation was a moderate 27 mm (Miomote Amedas), but peak intensity reached 5.5 mm/h. Fine weather melt pattern returns and is especially clear during 3-6 May with near-symmetrical sine waves.

The images below are taken from a time-lapse camera, showing the gauging site conditions (upstream view) during the rain-on-snow event from 24 April through to the peak stage of 0.894 m at 13:40 on 25 April. The colour of the water indicates the suspended sediment load, which increases rapidly as rainfall intensity increases after 0840 on 25 April.

The image below shows today's river condition (8 May). Fine weather melt continues and stage is 0.505 m. The gauging cross-section shows evidence of fine gravel accumulation which has caused a deviation from the established stage-rating curve.

Snowmelt runoff pattern emerges

The hydrograph above shows that the typical snowmelt pattern - daily fluctuations like a sine wave - become clear from about 14 April. This is because of rising temperatures and continuing fine weather.

Discharge measurement was undertaken with stage rising from 0.615m to 0.645m as daytime melt started to kick in around noon. We also collected data from a temperature sensor set in a cedar tree at 200m elevation. This will be compared to temperatures measured at the local climatological station near sea-level (Murakami Amedas).

Snowmelt on hold

A cold April so far, with fresh snowfall events even down at sea-level elevations. Today was only about 5 degrees Celsius, and turning to rain in the afternoon or snow in the mountains. The stage hydrograph barely shows the diurnal snowmelt pattern which is typical during fine weather in this season. Snowpack remains in some shady spots near the gauging station (see photo above), while the cedar lysimeter had melted off (see photo below).

Winter runoff data download

First time to download the water level data loggers after the long winter season. Here I am plotting the data from the Kadec logger, which shows there were three notable rain-on-snow (ROS) peaks (2 Feb, 20 Feb, and 12 Mar) all with a peak stage of about 0.8m. Low winter baseflow continued for about one month from late December to the first ROS event on 2 February. After the second ROS event on 20 February the base flow remains higher as we enter the snowmelt season.

Marginal snowpack remains in the forest around the stream gauge. The snow lysimeter at the larch site was relatively more exposed than at the cedar site. It is normal for the larch site to melt-off a few days before the cedar site. Melt-off dates seem likely to be close to average (end March to early April). The lack of ponding in the larch lysimeter trays indicates good drainage conditions with no problems.